Language Industry at a Crossroads – A Snapshot and a Daring Prediction

“To a man with only a hammer, every problem looks like a nail."

​Have you heard of the story of Samuel Brannan, California’s first millionaire, who made his fortune during the California Gold Rush in the 1850’s? When everyone was trying to dig for gold, he bought up all the mining supplies, picks, shovels, and pans, and sold them at a huge premium. He also spread the news of gold, which helped him sell more tools. This became known as the “pick-and-shovel strategy”, investing in the infrastructure of the boom, rather than the risky core activity. I am not sure if I read this story or if I watched a documentary about it – but I am sure that one way or another, this happened some 30 years ago. The story has stayed with me ever since. I forgot the guy’s name, but it was easy to find it for the purpose of this article. 
​
While we’re at the tools – and I’m going to refer to the proverbial hammer mentioned in the opening quote – you must have heard about the story of an expert engineer who was brought in to fix a critical machine that suddenly stopped working. He arrived, inspected the machine carefully, walked around it for a while, then took out a small hammer, tapped one specific spot once and the machine immediately started working again!
A few days later, the company received an invoice. Total: $10,000. They were outraged and said: “You barely did anything! You just hit it once!” So, they asked for an itemized bill. The engineer complied. “Hitting with a hammer: $1. Knowing where to hit: $9,999.”

Sometimes, I cry myself to sleep picturing myself as the protagonist sending righteous bills to my customers, €10,000 at a time, triumphantly making the point that it is my expertise that counts, not the time it took, and seeing them realize the bitter truth and accepting their fate.
From this point on in the article, I was about to segway into the language industry report, but one blog post by Seth Godin arrived in my inbox with an unexpectedly relevant theme for this whole topic. I will relay some parts of it that will make my point stronger. It is a story about a guy named Joseph Brandlin who was arrested for putting unauthorized stop signs at a neighborhood intersection he says has become increasingly dangerous for children, after the council failed to do anything about it despite his numerous attempts to warn them of the increasing danger traffic presented to children. Enters Seth:

“A hundred years ago, the default was that pedestrians were in charge. Cars were guests, only going where they were invited. But the persistent productivity and cultural force of the automobile carried the day, and the default flipped. The roads must roll.
If it can be paved or straightened or sped up, it is. If the car wants it, the answer is “yes.”
When systems gain momentum like this, it’s because they create urgent and immediate value, enough to disrupt the status quo. And once the status quo has changed, the momentum becomes normal […].
The system doesn’t care about Joseph Brandlin’s kid. It cares about the flow and the status of those that maintain that flow.
Every system eventually acts as if it’s more important than the people it was built to serve.”

And now to what I originally set out to write about.
This report is a mishmash of personal decades-long grudges and insights, sober historical perspectives, humble attempts at predicting the future, and insights shared by some of the most relevant voices in the language industry, which by itself requires a proper definition as it can be as vast or as narrow as we decide.

My definition will be a bit broader and include content production and localization (with translation, transcreation, adaptation at its core). It will only briefly mention interpreters. The reason is that interpreting often happens in higher stakes contexts or context that are at least perceived as involving higher stakes, and because it does not seem so lucrative to automate it. Also, you cannot fake your way into interpreting – the feedback is immediate and your future depends on your present performance.

Stakeholders

​This industry comprises several stakeholders: professionals (translators, editors, interpreters, copywriters), intermediaries (translation/localization/content creation agencies, marketing/advertising agencies), technology providers (generally, software companies providing tools to support the work of professionals and intermediaries), institutional stakeholders (professional bodies for translation and localization – mostly in the form of associations or institutions), job boards or similar platforms trying to connect professionals with intermediaries and consumers, localization departments within corporations, and consumers (private, corporate, public clients).
​
The second oldest industry in the world has worked mostly the same until about the time the internet spread globally, which brought about an inevitable transformation. The transformation was driven by technology, naturally, and by aggressive businessmen who saw an opening and swooped in full speed. The first big wave of transformation was driven by a widespread use of machine translation. Today, AI has touched almost every industry and it finally made it clear to everyone that what they spent their lives perfecting is now considered collateral damage of progress. With so many moving parts, uncertainties, daily technological updates that make yesterday’s knowledge and skills obsolete, societies cannot keep up nor survive. This is what Alvin Toffler calls “future shock”, a state of disorientation, stress, and overwhelm brought on by rapid technological and social change in too short a time. He introduced the concept in his prophetic and insights-packed book Future Shock, written some 50 years ago and predicting every single detail of everything that happened between then and now. Read it.

The Impact of Artificial Intelligence

As is the case with most knowledge workers and most industries nowadays, AI has dramatically shaken the language industry and may be the final blow that ends the industry as we know it. AI is affecting internal localization teams on the client side, and the whole industry on the supplier side, with language professionals leaving the industry in droves. Buyers expect better output, faster, and cheaper. This is nothing new, but in the last couple of years it might have gotten more attention. It must be mentioned, though, that on the premium side, both the suppliers and the brands are aware that localization is now an organic part of multilingual content processes. This is not enough to move the needle for the entire industry, though.
​
When it comes to high-cost and high-impact advertising agencies, they will survive only if they start handling more complex campaigns that involve digital, production, and integrated services, with many smaller ones eventually falling behind. In this space, perfection is giving way to rawer, more real content, micro-communities, micro-influencers.

An unregulated profession

As an unregulated profession, translation attracts wannabes, shady middlemen, quick-buck folks and pocket-money brigades, which creates confusion and distorts the market, resulting in downward price pressures, manipulation, and putting an equal sign between imposters and people who dedicated their lives to the craft.

The intermediaries (translation agencies, advertising/marketing agencies) have resorted to the heavy use of technology – delicious, I’d say – and aggressive freelancer “hiring” practices – no delicious at all, where freelancers have to go through onboarding that is equivalent to proper employment, with one significant difference: there is a clause they have to sign saying “nothing in this agreement constitutes employment relationship between Company and Contractor”. It looks like employment, it walks like employment, and it quacks like employment – yet it’s not employment. Cui bono? Or, translated from Latin: Who benefits?
​
Actual professionals doing intellectual work are thus stripped of any power and ownership of the final output. The only thing they are accountable for is everything. They do not own the process, yet any complaint from the agency’s client, whether grounded or not, trickles down to them and they end up being blackmailed into making corrections they should not be making in the first place or some other unreasonable request with a threat of non-payment or deduction of the agreed amount. I am not even referring to it as “fee”, because fee would be what professionals charge. The industry is dominated by “rates”. It is “standardized”, with processes being more important than the purpose, and with the only moving part being language professionals, who are “hired” and “fired” at will. This is perfectly illustrated by Jeff Bezos’ warning in his 2016 Letter to Amazon Shareholders:

“As companies get larger and more complex, there’s a tendency to manage to proxies. […] A common example is process as proxy. Good process serves you so you can serve customers. But if you’re not watchful, the process can become the thing.” – Jeff Bezos

“What do you think about AI and translation?

AI advances in language translation, content creation, and everything else you can imagine have been so great and so beneficial to people that all of a sudden they started wondering if anyone really needs professionals. I have had a fair amount of such questions from friends and acquaintances. I wish I had pulled them aside and said: “Look… your intuition is right. I don’t want to scare you but… We are done. I am done. The world as we know it is coming to an end. Let’s pretend it’s not happening to avoid panic but... It was good knowing you.” Experts can notice some inconsistencies and mistakes in the output that may escape an untrained eye. This argument has been used by many professionals to discredit all AI output, which does more damage than it helps. AI output is actually excellent to a high degree, sometimes unexpectedly good, sometimes better than expected, and it is definitely faster and cheaper.

So, where is the problem?

The above example follows the typical 80/20 rule. 80 percent of the output is excellent and 20 percent requires exceptional human insights, skills or knowledge. To develop the top 20 percent of knowledge and skills for the job, it takes 80 percent of your life or working life. One could argue that it is not worth spending 80% of your life to master 20% of the skill, when without any investment into skills you can have 80% of excellent output.
This reminds me of the famous quote attributed to U.S. retail magnate John Wanamaker (or to UK industrialist Lord Leverhulme): “Half my advertising spend is wasted; the trouble is, I don’t know which half.”

And that’s where the problem is!

You need an expert – be it a translator, a copywriter, an editor – to tell you if the output as a whole is good to go. The “20%” may be as little as 0.20% or even zero, but you need someone who is trained enough to tell you that – and to take full responsibility for it. If the market remains unregulated and if the army of bozos continues to drive professionals out by undercutting the prices, you will end up destroying a profession. People may love languages but no one will want to dedicate their life to a profession that does not provide for a dignified living. This is already happening. Universities face a lack of interest in their language or translation courses, which leads to fewer and fewer students, and will ultimately lead to no language professionals. There won’t be anyone left to “just take a look”.

"The biggest threat to translation quality today isn't AI itself, it's the illusion that AI output looks finished. Machine translation now produces text that passes a casual glance, which makes it harder than ever for clients to see why they need human expertise. The real value of professional translators was never in the 80% of text that's straightforward. It's in the 20% where cultural nuance, creative adaptation, and deep subject matter knowledge make or break a product. In game localization specifically, we're seeing this play out in real time: titles like Hollow Knight: Silksong and Elder Scrolls Online have had their Chinese launches destroyed by translation choices that no amount of technology could have caught, because the problem was never speed or cost, it was understanding." – Loek van Kooten, Owner at Loekalization

Adapt or die

Many voices are talking about adapting to the new reality, mastering new skills, going into post-AI translation editing, AI training, and similar – mostly neglecting the reality: these are not appealing jobs. Better yet, these are dumb jobs. Even if the pay matched what professionals used to earn, and it does not come even close, why would anyone want to do them? And do you understand that you are just helping a few big corporations replace you altogether?
​
I will never forget what Ildikó Horváth, Deputy Director-General for Translation at the European Commission, said a few years ago in the podcast I did with her: to future-proof their career, translators should learn how to translate. This is a real skill, and this is such a simple, yet deep message. I’ll take this advice any day over the “adapt or die” one.

We are at a crossroads

We are at a crossroads. It is obvious that AI and AI-driven language services have taken over and are winning, but the only market they are truly taking over is high‑volume, low‑risk translation or creative work. The do-it-yourself type of work. This is great news. I have always hoped for a day when we will not need translators, because everything will be translated perfectly into whatever language you choose. Working on that solution is a beautiful and a noble cause. However, I have to underline that our world was built by people and for people, and, just like Seth Godin mentioned, every system eventually acts as if it’s more important than the people it was built to serve. You know that song “Money makes the world go around”? Money, as one of the strongest incentives to do things, has done what an AI agent would do – see immediate gains, cut the most obvious costs, benefit the most invested persons. In doing so, it would neglect the most important person in the room, the creator, the interpreter, the human, because it ‘read’ that their work was not crucial. It was not protected in any way. Anyone could do it.
​
We, as a society, are doing away with one category whose importance we have not realized yet. We might realize it when it’s too late. But I want to pose a radical question, one that is as general as can be: Do we really need progress for progress’ sake? I would personally sign up for the ‘80s if time travel was possible and if the ‘80s time was forever. We humans do not have a common goal. Instead, we have billions of individual goals, but one comes up as a uniting goal: happiness. We can interpret it as tranquility as well. That may be a better goal.
Humanity does not seem to be ready for such inhumane disruption. Why should we make people feel inadequate, old-fashioned, downright stupid? Why so much stress? Cui bono?

Back to more practical things. I am a strong advocate of regulating the translation industry, protecting translators, editors, proofreaders, interpreters, copywriters (maybe copywriters need protection to a lesser degree, though), making this category exclusive in ways that pharmacists, notaries, or lawyers are protected. To your “why?”, I may ask “why not?” Exceptions have been made throughout history. Pharmacists at one point were no more than merchants. Theirs is a long struggle from being seen as traders of substances to being recognized as health professionals with specialized scientific knowledge and public responsibility. A similar fight can be put up to get recognition for the profession from traders of words to being recognized as language professionals with specialized subject-matter knowledge and public responsibility.

This means that, in addition to the existing formal education requirements, there should also be licensing systems (state exams and licenses), and restrictions on who could open and run a translation/language company. Newsflash: it should only be people who graduated from a recognized language/translation school and all language positions (including project management – especially project management) should only be taken by people who graduated from a recognized language/translation school. Additionally, project managers should be trained in languages they manage.

If nothing is done, the exodus will continue. There will be fewer and fewer people working in this industry as full-time professionals and there will be two classes forming: the crowd and the elite. The elite class has already claimed their position, but mostly due to some very specific characteristic they have. They are an isolated group that cannot really give advice because their position is very different from that of the bulk of the language professionals.

If you are still thinking what to do – to stay or to leave, there are two options only: if you stay, you can either accept the game, learn new things, work for less money, and be replaced eventually. Or you can stop being an accomplice to the demise of your profession, start acting like a professional, lose a lot of clients, suffer economically in the beginning, but regain dignity. Those professionals who stay long enough can only move upstream in the value chain: toward strategy, cultural nuance, and context. Theirs is the kingdom of God.

The language industry is not going away, and AI is not the only cause of its woes. It will change. It will have to change. The foundations for the future are made today by all of us, and understanding why we need to protect language professionals will lead us to how – by giving them the power.